Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Neural Correlates of Deep Comprehension of Second Language Reading: Focus on Retrieval Practice
Poster B26 in Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 4
Hyeonjeong Jeong1, Carlos Makoto Miyauchi1, Wataru Suzuki2, Ryuta Kawashima1, Motoaki Sugiura1; 1Tohoku University, Sendai, Japan, 2Miyagi University of Education, Sendai, Japan
According to McNamara (2007), comprehension can be categorized into "shallow comprehension" and "deep comprehension." Shallow comprehension involves understanding the superficial content of the language, while deep comprehension involves constructing new knowledge by integrating it with one's existing knowledge. Various factors influence language comprehension, including how the language content is accessed. Recent studies have reported that retrieval practice, which involves actively recalling information, promotes deeper understanding than mere reading activities (Karpicke and Blunt, 2011). However, the neural mechanisms responsible for such phenomena and the effects of using a second language (L2), which is less proficient than the native language (L1), remain unclear. In this study, using fMRI, we investigate the brain mechanisms underlying deep comprehension through ‘retrieval practice’ during the comprehension of L2 text content. We also examine the effects of language proficiency on these mechanisms by comparing L1 and L2 performance, as well as mere reading activities. The participants were 43 right-handed university students (20 females) with Japanese as their L1 and English as their L2. An English proficiency test (Oxford Placement Test) was conducted on the day of the experiment. Six texts on unfamiliar topics (politics, economics, music, psychology, science, etc.) were prepared in English and Japanese, selected from the TOEFL Reading section based on pilot tests. The stimuli texts were counterbalanced across conditions and languages to avoid reading the same content in both English and Japanese within participants. The text lengths were approximately 1,900 characters in Japanese and 450 words in English. Before entering the fMRI, participants read two texts in each language (L2 and L1) and understood the content under either the Retrieval Practice intervention, where they explained the text in their own words, or the Mere Reading condition, where they read the text twice. Inside the fMRI, parts of the texts from the Retrieval Practice, Mere Reading, and Novel Text conditions were presented randomly. Participants judge whether they had read the content before or if it was new while comprehending the content. After the fMRI measurement, a multiple-choice test was conducted on the texts read under the retrieval practice and reading conditions to measure individual comprehension levels. The results showed that behavioral comprehension levels were significantly higher in the Retrieval Practice condition than in the Mere Reading condition for both languages. For the brain analysis, using SPM12, we focused on the brain activity while reading the texts. The brain results showed significantly greater activity in the precuneus and left angular gyrus for the Retrieval Practice condition in L2 text comprehension compared to the Mere Reading condition. Furthermore, higher angular gyrus activity in the Retrieval Practice condition was associated with higher L2 proficiency, but no such effect was found in Mere Reading condition. These results indicate that retrieval practice promotes deep comprehension, particularly in L2 text reading, with the angular gyrus being deeply involved in this intervention.
Topic Areas: Reading, Multilingualism